Internalising 'Sensitivity': Vulnerability, Reflexivity and Death Research(ers)

AN0152008706;9eb01sep.21;2021Aug23.03:43;v2.2.500

Internalising 'sensitivity': vulnerability, reflexivity and death research(ers) 

Research about dying is viewed as inherently sensitive because of how death is perceived in many societies. Such framing assumes participants are 'vulnerable' and at risk of 'harm' from research. Simultaneously, with increasing recognition of the importance of reflexivity, researchers can become (deeply) preoccupied with their actions and experiences in the field. Whilst reflexivity is often described as a helpful process, in this paper we consider when introspection becomes problematic and even harmful for death researchers, in both a professional and personal sense. Identifying a process we call 'internalising sensitivity' the paper describes our own experiences of working and living with the pervasive ethical notions of sensitivity, vulnerability, risk and harm. We argue that these discourses can get 'under the skin' of researchers in that they impact researchers intellectually, emotionally, and physically, and this in turns affects their relationship with the research process and their place within, and beyond it.

Keywords: Ethics; vulnerability; reflexivity; sensitive research; death studies

Introduction

'Much has been said on what the researchers must do to avoid harming participants ... but less is written of the potential risk of harms for the researcher who studies sensitive topics and uses sensitive methods' (De Laine, [14], p. 75).

The starting point for this paper is that, whilst much has been written about conducting 'sensitive research' with some explicit focus on the topics of death and dying (Goodrum & Keys, [24]; Mckenzie et al., [44]), there is still limited discussion about the impact this has on the researcher (Burles, [9]; Visser, [58]; Woodthorpe, [60]). Here we unpack how framing death research as inherently sensitive, and therefore involving vulnerable participants and requiring the use of reflexive approaches, can lead to experiences of uncertainty, vulnerability, and harm for the researcher. By sharing our own experiences, we seek to expand theorisations of researcher vulnerability, beyond the primary discussions about personal safety and emotionality, to examine how the ethical discourses we operate within, and our reflexive attempts to navigate these, might contribute to our own vulnerability as researchers.

As scholars working in death-related contexts we recognise that our research is framed by other researchers, ethics committees and funders as inherently 'sensitive' (Addington-Hall, [2]). This very idea is deeply pervasive in encounters we have about our work. For instance, when we identify ourselves as death researchers, when we notice how similar research is discussed in academic spaces, and in particular when we have to navigate through the various concerns of ethics committees. Because of these assumptions, the participants in our research are considered de facto 'vulnerable' and at risk of potential harm. Whilst this is not uncommon in projects deemed to be 'sensitive' (Lee, [37]), it does not adequately represent how we view our research and its participants: these are not the ontological assumptions that we make about studying the end of life.

As social scientists, we appreciate how discourses are constructed and shape the realities in which we can operate (Hacking, [27]). Indeed, scholars have begun to challenge assumptions about 'sensitive research' arguing these can be reifying and stigmatising for participants, and (to a limited extent) researchers (Lee, [37]; Liamputtong, [40]). In this paper, we extend these critical discussions in a novel direction by examining how the discursive construction of sensitivity, which draws extensively on ethical notions of risk, vulnerability and harm, can affect the researcher, and in turn, the research they produce. The first section of the paper discusses this discursive context. Then, through reflexive analysis of personal examples, we show how these ideas are internalised by researchers and can feel weighty in intellectual, emotional, and embodied ways. We refer to these feelings as 'internalising sensitivity' and our examples explore the effects of this on academic and personal life. The final section problematises the assumption that reflexivity is inherently and always good. The wider salience of our observations for researchers doing 'sensitive' work and 'practicing' reflexivity both within and beyond death studies is also considered.

Background

'Sensitive' death studies and its 'vulnerable' participants

There are some topics that are presumed intrinsically 'sensitive' to study (Lee, [37]; Sieber & Stanely, [53]). Death in England (and in many other countries) is an obvious example. Although researchers have critiqued the label of 'sensitive research' for several decades now (Dickson-Swift et al., [18]; Lee, [37]), work related to death, dying, and end-of-life care is frequently categorised as sensitive in funding proposals, ethics applications, publications, and presentations. We suggest these generalised ideas about sensitivity derive in part from a negative, problems-based view of death and dying (Kellehear, [32]). Often when people talk about death research as sensitive, they do so in general terms rather than stating explicitly what makes it so. As Lee and Renzetti ([39]) note, the term 'sensitive topic' is frequently used in a common-sense way without being defined. When concerns are articulated they tend to centre upon death's emotionality (Cain, [11]; Cook & Bosley, [13]), or the fact that dying participants have limited, 'precious' time (Barnett, [3]) and reduced capacity to consent due to illness (Gysels et al., [26]). Death (and talking about it) is also considered 'taboo' in many countries (Lee, [38]). Even if sociological literature has problematized this idea (Kellehear, [31]), the label 'sensitive' may reflect generalised anxiety about a perceived social stigma associated with death. It points to an academic narrative about doing death research, which draws on societal discourses and expectations about what studying death 'is like', even if existing research literature speaks to a more nuanced way of understanding the field.

Moreover, it is not uncommon for the terms 'sensitive' and 'vulnerable' to be used interchangeably to describe death research, or other topics deemed to be 'sensitive' (Liamputtong, [40]). Since it is presumed this research is 'sensitive', there is a logical extension that its (human) subjects are essentially 'vulnerable'. For example, ethics committees are frequently concerned about how researchers will manage and respond to participants' emotions during interviews (Borgstrom, [7]). The most common presumption is that talking about death will upset participants; and whilst not unreasonable to have this concern, informal discussions with members of ethics committees indicate that similar fears may not always be extended with quite such apprehension to other kinds of 'sensitive' research topics. This vulnerability – or potential to cause 'harm' by initialising 'upsetting' conversations – is framed as a 'risk' inherent in this kind of research and something that responsible researchers are expected to mitigate and manage.

Vulnerability as problematic for research

Following the 'reflexive turn' in the social sciences, researcher emotions have, to some extent, become a focus of analysis (Parvez, [47]). Less attention has been paid to researcher vulnerability specifically and its methodological implications. This is despite those who research 'sensitive topics' feeling personally stigmatised by the 'taint' of their area of study (Hockey, [29]). Researchers can also be exposed to risks presented by their research methods (Blackman, [5]) and a burgeoning literature testifies that they experience a range of challenging emotions engendered by their work (Fincham et al., [22]; Hubbard et al., [30]; Woodby et al., [59]; Woodthorpe, [60]). It is now recognised that feelings can shape the data researchers generate, the course of their analysis (Mauthner & Doucet, [42]) and any subsequent conclusions they make (DeLuca & Maddox, [16]). Yet, often researchers are expected to conceal, deny, or demonstrate how they will minimise their vulnerability. Since consciousness of vulnerability can permeate the researcher's interactions with their participants, one's own vulnerability is considered an impediment to access and data collection (De Laine, [14]). Vulnerability of both participants and of researchers is therefore often framed in the research literature as problematic for the wider research endeavour as well as for individuals. Even in disciplines which seek to foreground human experience, acknowledging and exposing the vulnerability of the researcher can be difficult and problematic for the research process, and one's career (Behar, [4]; Visser, [58]).

In procedural ethics (Guillemin & Gillam, [25]), the vulnerability of participants and the safety (rarely 'vulnerability') of researchers is also deemed problematic by committees. The committees and their documentary processes are essentially concerned with identifying risk – the likelihood and extremity of which is considered greater in 'sensitive' contexts where participants are understood to be more vulnerable. Thus, the underpinning rationale for ethical governance is the procedural avoidance of harm caused by research studies and indeed, by extension, the people carrying out this work. Its association with harm and also exploitation means that vulnerability is largely considered as a negative thing and something to be avoided, mitigated and managed.

By means of its processes of formalisation, procedural ethics also gives the impression that issues like 'sensitivity' and 'vulnerability' can be rendered 'knowable'. It implies that many potential risks, harms and uncertainties can be managed in linear, foreseeable ways via the spelling out of protocol, delineation of categories and rarefication of the research process (Van den Hoonaard & Hamilton, [57]). Although some researchers have been critical of the excessive and inappropriate application of notions of 'harm' from biomedical science to research in the social sciences (Burr & Reynolds, [10]; Dingwall, [19]; Haggerty, [28]), these ideas have been exponentially pervasive and have significantly shifted attitudes towards the regulation of all research involving human participants. This places a weight of responsibility on researchers to be able to operationalise discourses such as sensitivity – to turn these ideas into knowable quantities, situations and scenarios in the context of their proposed work. They are required to make (convincing) plans for how they will eliminate or at least significantly reduce the likelihood that 'bad', unethical things will happen. By extension, this process can feel like a personal measure of the robustness of one's own moral sensibilities. Frequently within death studies, reflexivity is cited as a methodological tool that can be deployed for negotiating these issues (Visser, [58]). And whilst this has led to a useful recognition of the possibilities for disjuncture between procedural ethics and situated 'ethics in practice', it has also expanded the meaning of reflexivity from a largely methodological concern to one that also carries important ethical responsibilities (Guillemin & Gillam, [25]).

Reflexivity as personal responsibility

In the social sciences, reflexivity – an explicit self-analysis of one's own role in research – has become an expected element of qualitative research. This is particularly the case in contemporary ethnographic work where the researcher can forge prolonged and close relationships with participants. It is now generally understood that through careful reflection we can start to unpick the tangled subjectivities of researcher and researched and to recognise how one's power, positionality and biography shape the research process and the knowledge we produce (Dean, [15]). Thus, the fundamental relationship involved in the pursuit of reflexivity is one between thought (reflection) and action (reflexive practice) and this has its origins in the philosophy of individualism (May & Perry, [43]). In the case of 'sensitive' research, reflexivity is often presented as a responsible means of negotiating issues such as vulnerability and exploitation, which essentially involves the self-management of individual researchers and their potential to 'do harm' via diligent introspection.

Denzin ([17]) has suggested that the 'ideal' researcher is ' ... a morally involved, self-aware, self-reflexive and interacting individual who holds the self personally responsible for the political and ethical consequences of their actions' (Denzin, [17], p. 277). Whilst important, contemporary researchers are striving to attain this ideal and to meet their ethical responsibilities within a complicated regulatory context proliferating with guidelines, policies and procedures, which direct and delineate what 'good' (correct) research looks like. In our experience, this can feel overwhelming for 'junior' researchers in particular. And so whilst we agree with the many accounts which argue that reflexivity is important – that it is methodologically helpful and constitutes ethical practice (Guillemin & Gillam, [25]) – we acknowledge its problematic elements (Riley et al., [50]) and highlight the neglected issue of the various 'costs' of introspection for researchers.

In the following section we 'trouble' our own reflexive experiences. Through writing about our challenges, we raise the possibility that cycles of introspection about the 'sensitivity' of one's work and what this assumes about a potential to act unethically, exploit vulnerability and cause harm, may ironically engender vulnerabilities in us as researchers.

About the research

Our 'sensitive' research projects

Over the last 10 years, we have both conducted ethnographic fieldwork in England about care and relationships towards the end of life. This section describes the projects that gave us the specific experiences we draw on within this paper. Whilst we focus here on two studies carried out at the start of our careers as death researchers, our subsequent research experiences looking at multidisciplinary team working in healthcare settings, online forums for individuals with life-threatening illness, and baby loss and post-mortem, also inform what we have come to describe as 'internalising sensitivity'. Like Smart ([56]), we consider sustained involvement in qualitative research as biographically significant for the researcher and our research careers to date foster an understanding of the 'fuzziness' of research beginning and endings. In this sense we value Smart's ([56]) description of data as research fragments that accumulate over time, acquiring the potential to reveal, affect and haunt us in different ways.

'Constructing' end-of-life care

Erica's research has focused on how the 'end of life' is constructed, practiced and experienced, from the perspective of policy(-makers), those providing professional and informal care, and those who are supposed to be the subject of end of life care. Her largest fieldwork to date on this subject involved observing 50 hours of policy-related events, 250 hours of participant-observation in clinical settings, care homes, and disease-related support groups, and 100 interviews with policy-makers, health and social care professionals, and people living with life-limiting illness and/or their partners/family/carers. She also spent up to 14 months regularly visiting people living with life-limiting conditions near the end of life (and their families and health or social care professionals were possible) in their homes and other locations related to their everyday life and care (Borgstrom, [6]). These visits included anything from informal chats, to sharing family meals, to accompanying people on their clinic appointments or visiting them when in the hospital and ranged in the spectrum of participant-observation depending on the context and arrangement with participants. Encounters included witnessing emotional and/or physical distress, sensing family tension, and in some cases, being present shortly before the death of the person. As part of the research design process and ethics committee approval, these kinds of scenarios and 'appropriate' responses were explored, including how to ensure participant consent, provide for researcher safety, and account for the potentially emotionally charged nature of the research. Throughout the research, parallel field notes and field diaries were maintained, and combined with regular academic and peer-support supervisions. A reflexive approach was inherent within the project's methodology and was actively used to frame the study's findings. This is evident in the doctoral thesis, where extracts from the field diary were used to introduce each analytical topic.

Family practices and the end of life

The experiences Julie describes in this paper relate to her ethnographic work exploring how everyday family life is pursued when someone in the family has a life-threatening or terminal illness (Ellis, [20]). This study used the concept of family practices (Morgan, [45]) to theorise the mundane and everyday aspects of living as a family nearing the end of life both at home and when staying on a hospice inpatient ward. Julie completed 175 hours of participant-observation at the ward where she took-up the role of hospice volunteer, chatting with patients and relatives, making drinks, tidying around and assisting at mealtimes. To record her reflections, Julie kept fieldnotes which integrated observational details, analytical thoughts and emotional and personal responses to spending time with severely ill people and their family members. She also conducted 39 in-depth interviews with relatives from nine different families where one member was living with a life-threatening illness. These were very informal and involved repeat interviews with some family members where circumstances and illness trajectories allowed. Interview participants were recruited via the hospice's day care service which is based in the North of England. Fieldnotes were completed following each interview encounter, recording important contextual information and the researcher's personal feelings, which were reflected upon during analysis. Both interview transcripts and fieldnotes were coded and analysed using a narrative approach to thematic analysis. Like Erica, a reflexive methodological approach was adopted throughout the research process. It featured regularly during formal supervision with Julie's doctoral supervisor. At the time, the personal toll of completing the work was discussed and vulnerabilities, which feature in the example below, could be evoked, although they never felt fully articulated.

Our experiences of internalising sensitivity

We have chosen the following examples after careful consideration and discussion about how the discourses of 'sensitivity' have impacted upon how we experience our research and how they continue to affect and haunt us in different ways (Smart, [56]). We have shared similar examples previously in conference presentations (with varying receptions), in discussions with colleagues, and in our PhD theses and subsequent publications (Borgstrom & Ellis, [8]). However, some of the specific instances we discuss here are making a tentative and somewhat nervous first appearance in print for a wider audience. Tentative because we do not want to suggest that the examples offer comprehensive coverage of the issues we explore, nor do we expect they are particular to us. We offer them here as an invitation to others to share their experiences of living with 'sensitivity' in the various contexts of their work. We are also very aware of how we may have selected certain examples rather than others in an effort to 'protect' ourselves as 'vulnerable researchers' from the risky and exposing nature of reflexive publication (Blackman, [5]; Dean, [15]). With these issues duly acknowledged, the purpose of us offering these examples is to enliven methodological discussion. In doing so, we hope to expand theorisations of researcher vulnerability that extend beyond issues of personal safety and emotionality. This enables us to examine how the ethical discourses we operate within might contribute to our own vulnerability as researchers.

We have selected three examples:

Experience of family bereavement immediately after fieldwork involving end-of-life care

Being alone with a dying person during fieldwork

Writing about 'the mundane' towards the end of life

In each of these sections, we write in first-person about our accounts and experiences, sometimes drawing on adapted extracts from our field diaries to demonstrate the visceral and emotional sense of vulnerability as it 'got under our skin'.

Can I just feel it? Family deaths filtered

Here we describe how discourses around death studies shape how we view ourselves and our personal responses to death outside periods of formal fieldwork. As ethnographers, we are trained to think about our own everyday experiences as on a continuum with fieldwork, a continual quest to make the familiar strange. This has enabled us to note and seek to make sense of these occurrences within a framework of 'doing' research. It has also, perhaps, desensitised us to how strange it may be that the discourses that surround the type of work we do have affected our 'non-professional' lives. To illustrate this, Erica reflects on caring for her paternal grandmother, an experience which highlighted the impact of her exposure to the idea that participants in death research are inherently vulnerable.

Weeks after finishing my fieldwork about end-of-life care in England, I was back in the USA visiting family. Whilst I was there, it was revealed that my grandmother had recently been diagnosed with cancer. She was in the midst of deciding to decline further investigations and treatment. Soon I would visit again and effectively become her 'carer'.

During this time, I had a desire to separate work from personal life, and an inability to do so. Months later, my mother remarked that my grandmother was excited that I could be 'studying her', something I resisted in practice. I wanted to just experience the moments for what they were, not to be making mental fieldnotes. I wanted to have memories that were not categorised or tainted with anthropological theories, analysis, or attempts to generalise from the particular.

Yet, it was when I was caring for her, doing the gardening one afternoon that I really connected with my previous fieldwork experiences. As I picked up sticks from the grass, something reverberated through me that made me recall a conversation I had with a man who knew he was dying, a side remark about how much waiting there is in the process of dying. I could begin to understand aspects of what my participants told me that seemed trivial before.

It was only later that I realised that my desire to separate my personal experience from my fieldwork mode was because I thought I needed to protect something, and someone – me and my experiences of personal deaths. I had, through the way research is framed in death studies and in healthcare research, seen myself, as a soon-to-be-bereaved grand-daughter, as someone who needed to be protected around the time of death. Someone for whom intrusion – from research – had to be limited. I had projected myself as a 'vulnerable person' because that is how the participants of my research are often portrayed; and I was now in similar circumstance. But without this notion of vulnerability, I did not necessarily feel I was particularly emotionally vulnerable because of being a carer to a dying grandmother. It was because I thought I should be vulnerable that caused me to become unsettled.

Moreover, I began to see myself as 'uniquely vulnerable'. Vulnerable both because of the situation (caring for a dying grandmother), but also because of the heightened knowledge I had about death due to my research. I had gained knowledge about death and dying that most people don't have until they experience it first-hand. And I became angered because there was no one to protect me (the grand-daughter) from me (the researcher with knowledge), and I blamed reflexivity for my ability to be critically aware of this dual positionality. Those who knew both my research and personal situations did not view me as 'vulnerable'. In fact, the opposite was often presumed: that if I could 'handle it' professionally, then personal encounters with death would not be particularly difficult.

Reflexivity, although considered a prerequisite of contemporary social science research, can be quite unpredictable (Kress & Frazier-Booth, [35]), and this series of realisations felt unpredictable. This example highlights how the discourses of vulnerability and sensitive research that permeate death studies can impact our own personal experiences of death. It also illustrates how being reflexive can heighten this awareness, without necessarily providing an outlet for the emotionality that it engenders. In other contexts, this dual positionality is viewed as potentially productive; for example, in creating more compassionate doctors when the doctor self-identifies as a patient-practitioner (Rowland & Kuper, [51]). Whilst the dual positionality here may have furthered aspects of Erica's analysis, it also had unintended consequences of disturbing the experiences of familial death.

Should I be here? What to do when observing

A common experience in our fieldwork in end-of-life contexts is asking ourselves: should I be here? Should I be witnessing this? On and off throughout our careers we have been fearful of the charge of voyeurism (Hockey, [29]) and what our desire to observe 'vulnerability' in this rather 'extreme' sensitive context might suggest about us as 'moral' individuals. Here we provide a particular example from Erica's research where her internal dialogue about whether this was the 'right' thing to do manifested in deeply embodied ways in the field (Okley, [46]). What Erica notes echoes experiences that Julie had during her fieldwork when she agonised over if and when it was OK to enter a patient's room on the ward. This was experienced in embodied terms in her discomfort at hanging conspicuously and awkwardly around door frames.

I found Mable in the care home. She had been discharged after her rather lengthy, unexpected hospital stay. A care worker showed me to her dowdy room, where she promptly left me. As I peered around the doorframe, I saw Mable in the bed. I softly announced my arrival and made my way to the room's only chair. Sitting down, I explained to Mable what I was doing whilst simultaneously unsure of myself. Mable had previously said I could visit, even if I found her asleep, but this being the first time, and her body restless, it felt odd to be there. I wasn't quite sure how she was doing or what this sleep 'meant' – could she be in her last days of life?

I took out my notebook, like I felt a dutiful anthropologist should, and jotted a few lines about the layout of the room and what was going on. The chair I was sat on was at the foot of her bed. I realised I couldn't just sit and observe her. I was reminded how the university research office joked about how tabloid papers might report on my project: 'researcher watched my mother die'. They were worried not just about the ethics of the project, but the social and moral response to such a 'sensitive topic'. But, at the same time, I was constrained by 'ethical procedures' to 'not doing anything' – according to the ethics committee and university research office, I was to observe, not to intervene. Yet, what would 'intervening' in this situation look like?

At some point during my visit, I sat on Mable's bed near her legs. I continued to speak softly to her, talking about the flowers and saying things like 'it is okay'. I did not want to wake her. Thinking back, I do not know where those words came from. I lightly touched her, and I recoiled. Not because I did not want to offer the compassion that I associated with the action, but because I questioned it in light of the 'ethics' of doing death research. What right did I have to touch her? What obligation did I have not to? When I visited Mable in the hospital, she often wanted to hold my hand as I listened to her stories. At the time of writing this article, I can still feel the folds of her paper-thin skin in between my fingers as I type.

Recalling such scenarios serves to illustrate how discourses around the sensitive and vulnerable nature of doing research about the end of life makes us question these actions in the field. The actual details of what it is like to be alone in the presence of someone who, as it turned out, was in their last few hours of life was not something that was discussed in all the research design and ethical approval processes. Procedural measures covered consent, not how it would feel. The other times Erica had such 'exposures' was in the company of nurses or doctors, serving both a potential model for behaviour but also an additional site of research focus. Being alone with Mable, Erica was left to wonder what she should do whilst in the room and even if it was appropriate for her to be there at all. Was Mable 'too vulnerable' for Erica to be witnessing her toss and turn? And this, in turn, clashed with the rhetoric about 'not being alone' towards the end of life as being an integral part of a good death (Seale, [52]). Without Erica present, Mable would have been alone, and yet at the same time, Erica felt internally as if she herself was deserted. Left without a guide in how to be. It is an example of how even with detailed ideas about how fieldwork will be conducted, it is not always possible to foresee all encounters (Cohen, [12]). This lack was apparent to Erica in how uncomfortable she was in her body whilst in that room. The seat was not quite right – she could not settle for long, as if she was unconsciously mimicking how Mable tossed in the bed. Sitting on her bed felt simultaneously wrong and right.

What this experience also highlights is the concern that researchers – and those who supervise or fund their research – have about the stigma attached to conducting 'sensitive research' and how their methods may be interpreted (Lee, [37]). Reflexivity has enabled us to acknowledge that this fear of stigma influenced the research office's remarks about the project, and in turn affected how Erica made sense of her own actions on the day. Nonetheless, reflexivity cannot mitigate the feelings of stigma, discomfort, and shame which persist long after the fieldwork has ended and continue to make writing about such experiences feel risky (Dean, [15]).

Can I publish this? Data and revealing

In this final example, Julie shares her thoughts about 'the analytical aftermath' of doing research, acknowledging that the post-study period is not readily discussed in terms of how it affects the researcher and the emotional labour involved. Here we focus on an example of 'chilling'. Chilling is when researchers defer or deliberate about dissemination of research on a particular topic because they await possible hostile reactions (Lee, [37]). The following are notes from Julie's fieldwork diary taken at the time of the study.

In the car driving home, I felt strange - subdued and numb, but also a little tearful. It made me sad to think about what Malcolm and Tracey[1] are facing and I began to feel troubled about putting the thesis together. I get mixed feelings - sometimes feeling bad for having a numb and detached feeling and not being overwhelmingly burdened by the circumstances I find people facing. Perhaps their ability to just get on and do makes me able to be saddened but not paralysed by distress. But then, I do feel compelled lately to reflect on my own life and the time I have and how I use it ... time is passing quickly and studying dying is magnifying the importance of maximising it. I feel so guilty because I imagine I ought to be more committed to the thesis and not have to struggle to sit down and work on it. I'm aware that the sheer enormity of the task is overwhelming - I feel lost for ideas - not sure of what I want to say about any of this. Could this explain the numbness and detachment - the lack of motivation? Of course, feeling indebted to the families who have spoken with me is an added burden - I feel like I'm struggling with getting themes to emerge - I'm 'feeling' the 'ordinary' yet specific life worlds of individual families and I am starting to wonder if illness experiences within families are just different and what more can I say than that? If I simply re-tell their stories how am I doing anything they couldn't? Ultimately the work seems trivial ... imagining Malcolm's isolation [due to deafness] makes the thesis feel small, like nothing - like it couldn't do anything to touch the magnitude of what that experience must be like - yet the mundane comes in as the everyday world keeps on ticking over. What can I say about this?

As this extract shows, Julie was troubled by any 'mundanity' in her own approach to, and reflections about, the research. Whilst this suggests a sense of ambivalence about generally being able to get on and not feel emotionally overwhelmed, it also points to difficulties in analytically working out how to make sense of the 'everydayness' in the data. In the process of considering her own emotions, she began to see how she was struggling to formulate an analytical perspective on how the everyday and the mundane intersected with the more extraordinary and emotional experiences associated with dying. Assumptions regarding the intensity and drama of death meant that constructing an account that gave voice to the mundane and everyday present in family stories was challenging. This was both analytically and emotionally challenging as Julie felt in danger of trivialising such a 'momentous' issue.

Against a background of discursive framing which tends to conceptualise dying as fearful, denied and ontologically troubling, it has felt quite challenging and exposing to talk about interpretations of data which foreground the 'everydayness' of dying people's feelings and the ordinariness in their accounts. Whilst it has not been Julie's intention to suggest that living with life-threatening illness is not an extraordinary, emotional and difficult experience, there is a worry that in making a case for the explanatory relevance of the mundane, it might be seen this way by others. Whilst even the most robust researchers might feel anxious about whether the patterns they see in their data and the interpretations they present as 'findings' are accurate, fair and make sense, these worries and crises of confidence feel especially difficult to shake when researching 'emotive' subjects like dying. In other words, discourses around participant vulnerability have heightened our own vulnerability – induced by an ever-present but unhelpfully nebulous awareness of our potential to cause 'harm' – something we carried throughout fieldwork and which persists into the dissemination. Being reflexive has helped us to be attuned to this potential to cause harm but it has also, particularly in Julie's case, encouraged a level of critical introspection that at times felt almost destructive (personally and professionally). For her it has not always been possible to retain perspective on the limits of what one can feasibly control – particularly at the point of dissemination when control must be given up to the scrutiny of others. And so, with this last example we illustrated how researcher vulnerabilities can become entangled with and fuelled by a concern to always act 'sensitively' and this can have a direct impact on managing aspects of the research process.

Discussion

In this paper we have drawn on reflexive examples from our own research to illustrate how discourses of vulnerability, due to our research being 'sensitive', have been internalised and 'gotten under our skin'. These examples illustrate how this internalisation has influenced the way we feel and think about our research and personal experiences. Our discussion of reflexivity illustrates how 'fragments' of research often extend beyond the boundaries and time limits of individual projects, and can continue to 'haunt' us (Smart, [56]). In attending to our own vulnerability, we see the possible limits of current ethical framings of 'sensitive research', and of reflexivity as a methodological tool, where our points are aligned with discussions about the need to do ethics-in-practice rather than be procedural (Pollock, [49]; Sikic Micanovic et al., [54]). Beyond acknowledging this, sharing these examples generates a space where the limits of reflexivity within sensitive research can be explored.

Firstly, we are not suggesting that reflexivity is not useful – it is often very helpful. We both continue to work reflexively; this paper is an example of that. Reflexivity has enabled us to make analytical connections through introspection. For example, Julie was able to wrestle with what appeared to be the paradox of the mundane in the extraordinary. And Erica began to understand what participants had told her after going through similar experiences herself, a phenomenon familiar to other anthropologists who have experienced a personal bereavement (Behar, [4]). Reflexivity also enabled us to question our positionality in the field and when collecting data, as illustrated in the second example. This is often the reason cited for why reflexivity is utilised within the social sciences and related disciplines (Abdullah, [1]; Kleinsasser, [33]). However, as evident in ethical discourses about protecting 'vulnerable participants' of 'sensitive research', reflexivity tends to be associated with the idea of protecting participants from harm (Kumar and Cavallaro, [36]).

Reflexivity is used within social research to mitigate (primarily) participant vulnerability and/or harm caused to participants through their engagement with research. The individual researcher is expected to use a reflexive approach to consider power dynamics, biases, and subjectivities (Pillow, [48]). Therefore the researcher – as a person – is an instrument in the process (Geertz, [23]), and the site for mitigating vulnerability and harm. One 'solution' offered for managing sensitive research, therefore, is to manage the individual researcher via introspection and reflexivity. What has been less explored is how for the individual researcher, this responsibility may be challenging and potentially harmful.

By drawing on examples that highlighted our own vulnerability, we raise awareness about the limitations of reflexivity in terms of protecting researchers from 'harm'. We have done this by sharing how reflexive practices have amplified our own sense of vulnerability at different times in our lives. While as a research community we are getting better about discussing the emotional implications for the researcher of doing sensitive research, attention is generally on the period when researchers are in the field engaging with participants (Dickson-Swift et al., [18]; Woodthorpe, [61]). Other aspects of the research process – such as producing notes and transcripts, performing detailed analysis, recounting stories for dissemination, or even listening to other researchers as part of peer-support – are also emotionally tiring and pose moral dilemmas (e.g. Evans et al., [21]). It is important to note that often these experiences can be highly individualised, personal and lonely, even in team environments because whilst research teams may sometimes write ethics applications collaboratively, the responsibility for translating these linear narratives into actions (or inactions) in the field lies with individual researchers. At times, as this paper demonstrates, doing research and living with its consequences has felt as if it could be 'harmful' to us as people. That we have begun to consider that we could be harmed through the research and reflexive process further indicates the internalising of the vulnerability discourse.

This sense of our own vulnerability resonates with Smart's ([56]) suggestion that social researchers can be 'haunted' by living with the fragments of other people's lives. For us, experiencing aspects of our research as 'haunting' is rather apt, given that data analysis actually involves 'working with the ghosts' of dead participants (Komaromy, [34]), which have accumulated through our various projects generating a cumulative effect over time. Hearing their voices in recorded material or recalling time spent with them, makes the responsibility of producing an account of their experiences even more overwhelming. Smart ([55]) suggests these feelings of responsibility are central to developing a 'sociological conscience': the intersection of a researcher's personal life with the practice of doing research. This haunting collides then with an ethical and moral imperative for us to not only produce 'good data' (Kleinsasser, [33]) but to be also ultimately 'good researchers' (Macfarlane, [41]). However, to-date, there appears to be little space within the academic community to comfortably share these experiences.

When writing this paper, we have reflected on the practices that have helped and hindered our experiences of vulnerability. Individually and collectively we have spent a considerable amount of time thinking, writing, and talking about our experiences, benefiting from doing this together and with other colleagues and friends. What has helped us beyond the reflexivity, and we understand may not work for all other researchers, are several things around the structures of conducting sensitive research. Firstly, pacing data collection (e.g. only conducting one interview in a day or only three in a week) and analysis can provide time to emotionally process the encounters as well as manage the energy required to be present during sensitive research. Whilst we experienced more control over timings during our doctoral studies, we are aware that due to time and/or monetary constraints of some funded research projects these can have periods of intense data collection and/or analysis. We recommend, where possible, those responsible for designing projects purposefully and carefully consider the timing and pacing of the various aspects of work in terms of the impact it has on those in the team primarily responsible for carrying out data collection and analysis. In doing so, it becomes less of an individual responsibility to manage the emotional and temporal intensities of the project, whilst at the same time, this can create an ethos within the project team where emotional labour can be recognised and discussed.

Secondly, the way project proposals (and grant applications) may be written can presume that research assistants/fellows are hired in to conduct sensitive research and focus solely on a particular project. However, such approaches can formally minimise the recognition that these are individuals with research (and personal) biographies that will influence their experiences on any one project. Whilst this is sometimes addressed when researchers openly acknowledge their positionality and/or discuss how their analysis was informed by previous work, we have found it helpful being able to engage with and think across our past projects and other entanglements in sensitive research. Whilst we have managed this informally, and sometimes with non-academics working in related fields who can provide different perspectives or through therapy, we would welcome research cultures that engage with the legacy of one's research in terms other than 'research impact' and career progression.

This relates to our last point, which is about the expectations of 'timely' writing and dissemination. We have found that sometimes waiting to write about a project or particular elements of it can provide us with new perspectives because of the distance it provides. Or, as in this paper, more confidence to reveal aspects that at times feel very raw and emotionally revealing that is counter to most of our professional cultures. Yet such delays can cause problems with meeting institutional and project objectives or activities such as the Research Excellence Framework. More could be done within institutional and funding structures to accommodate publishing patterns that account for researchers' experiences, wellbeing and reflexive practices.

We are therefore adding a note of caution to the idea that reflexivity can always function to alleviate vulnerability in sensitive research, particularly when experienced by researchers. Specifically, what reflexivity does not always provide is a solution to the 'living with' of our research and the effects of the vulnerability it can engender. By introducing the idea of 'internalising' discourses such as sensitivity and vulnerability we hope to offer a conceptual means for researchers to talk and write about similar difficult feelings in a way that avoids moral judication of the research or the researcher. Whilst our experiences stem from our work in death studies, what we describe has broader implications. It is particularly useful for the growing field of sensitive research, and thinking about the longer-term impacts of conducting research that is 'sensitive'. Being able to discuss how research experiences can indicate internalised sensitivity is useful in enabling us to think critically about research methods, training of researchers, and communities of practice.

Acknowledgments

The authors would like to thank the people who have all been part of their studies and those that have supported them along the way. Additionally, the authors would like to thank the editors and reviewers for their feedback on earlier drafts.

Disclosure statement

No potential conflict of interest was reported by the authors.

By Erica Borgstrom and Julie Ellis

Reported by Author; Author

Dr Erica Borgstrom is a Lecturer in Medical Anthropology and End-of-life Care at the Open University. She has published on a variety of issues relating to end-of-life care and death studies, and has co-edited with Dr Julie Ellis and Dr Kate Woodthorpe a book about researching death, dying and bereavement. She is co-editor for Mortality and sits on the council of the Association for the Study of Death and Society. Dr Julie Ellis is a Senior Lecturer in Sociology of Health and Illness at the University of Huddersfield. Her research interests include personal relationships, death, dying and bereavement, materiality and the everydayness of illness experience. She was recently the co-convenor of the British Sociological Association's Social Aspects of Death, Dying and Bereavement study group and is on the editorial board for the journal Mortality.